Guitar Amp Heads - Making a WEM More Marshall?

[rodabod]

I hope you guys don't mind guitar amp stuff here too much. I thought I'd be best asking here as there is no bullshit, and a lot of clever people.

Anyway, I have a Wem Dominator 15W valve amp which I use as my bedroom amp. I can't play it loud which is a shame because it stays fairly clean even at neighbour-annoying levels.

So, I am considering modding it to get some preamp distortion of the Marshall flavour. I have a DSL and have used JMP/JCM800 master volumes which I like.

Funnily enough, the amp is very similar to the Marshall 18W since Wem sold the basic circuit design to Marshall. I think Marshall probably modded the design to make it rock though...

The Wem schematic is here:

http://www.vintagehofner.co.uk/britamps/watkins/schematics/dommk3.html

and you can see the similarity it has to the Marshall schematic here:

http://elektro.cmhnet.org/~graydon/photos/musicstuff/marshall/marshall18wattrev4.pdf


So, for starters, I suppose I could just mod one channel of my Wem to that of the Marshall 18W preamp and see how it goes, but even then, it doesn't look like I'm not going to have much distortion at low volumes. I'm not sure though - I haven't tried one, but the lack of a preamp gain control doesn't suggest it would provide much.

As an alternative, I was thinking of changing part of the preamp to that of a early JCM800 as seen here:

http://www.tone-lizard.com/images/JCM800_pre.gif

In this case, I would maybe copy the first two stages and feed what is connecting to the third triode in the schematic (to the right of #3) into the master volume pot of the original Wem schematic.

I'm not sure if this suggestion is complete madness. For starters, I don't know if those two triode stages which I would have copied would provide much distortion, and also I am not very savvy with valves and their corresponding input and output impedances when in circuit. Also, I only have space for three knobs and a bright switch which I suppose means preamp gain, master volume and "tone".

Any suggestions would be appreciated. All I want is a good Marshall-style distorted sound - I appreciate that the output section will still remain (it's good anyway). Also, since the amp has two channels, I can keep purists happy by maintaining an original channel.

Cheers,

Roddy

 

[Gus]

looks like the marshall 18watt has less preamp gain, having only one triode then the tone then the PI.

The wem has a gain stage then tone stack then triode gain stage

the PI setups are close 47K 1k 56K 820

Now I would make sure what I did was reversible Is the WEM worth anything?

First thing to try is a different speaker something like a greenback or even an old TV console woofer etc. A speaker that "breaks up" at low power. One cool amp I heard was an old painted black tweed SE 5 watt princeton with a 12" TV speaker. Stock circuit

Speakers and cab are very important

Next try playing with adding and or changing bypass caps at the resistor cathodes in the preamp .68uf 2.2 uf films and 33uf electros are good starting points. The order of the gain and eq set by the cathode caps can matter.

Third "Revoice" the EQ. Duncan's tone stack sim program can help. Maybe make it a marshall cap selection in the tone stack.

 

[rodabod]

Hi Gus,

Yes, I did wonder about the Marshall 18W as it didn't look as if it had much gain either. Probably not enough for me!

The Wem probably is worth a bit (maybe a few hundred pounds), but I'm not too bothered about modding it. I found it next to a bin in the street.

Funnily enough, it already has a a 15" Greenback. It's a great speaker - the low-end is fantastic. The cab is large and open-back and sounds good to my ears. I'm not sure how much I like speaker distortion actually.

As far as I understand, the cathode cap acts like an emitter bypass cap in a transistor circuit, so you get an increase in AC gain while maintaining the DC conditions. So, if I start adding cathode caps, I'll get more gain at the point where the reactance of the cap is low (ie. higher gain at higher frequencies). So, are there any general rules I should apply, or can I just try different cathode caps like you suggest to see how it sounds? What I would ideally like is a lot of gain like you can get on a modern marshall, but without sounding crap!

Cheers,

Roddy

 

[dabo]

Choose the channel to modify before you begin. One channel has plate bypass caps and one does not. Keep the one you like best as it is. If you've decided to keep the one without these caps, remove them from the one you're modifying.

Reduce the input resistor to 33k or lower.

Try a cathode bypass cap on the first gain stage as Gus suggested. I recommend 22uF as a maximum value. Some folks like less or none. (Be aware that the gain increase mentioned will only happen to a point and beyond that you'll just be over-bypassed.) The coupling caps in this amp are mostly .01uF so the increase in bottom will probably be ok. If it gets mushy and flatulent reduce the cap value.

Tone stack revoicing is a good call too. The Duncan calculator is great for seeing how the values interact. Be aware that the Marshall voicing is related to not only cap values but how they scale relative to the pot values. The WEM tone stack is closer to a Fender in this regard.
I'd begin by reducing the .1uF to .047uF or .033 uF. (Check the calculator to see how it effects the bottom extension!) Controlling the low content is crucial to amp voicing (my opinion stated as though it were fact). Voicing the amp to the speaker is important (as Gus points out).
Listen at each step. If you disagree with me, you're opinion is the one that is correct. It is, after all, your amp.

Since there is a volume pot after the second gain stage you may get away with increasing the 100K plate load resistor on the first stage to 120K or 150K and hit the second stage a bit harder. This can be a dangerous game if you get too gain greedy because the bias point will move and you really should also rethink the cathode resistor value and the bypass cap value too. You may end up with really bad sounding distortion. Then again, you may enjoy the flavor of this really bad sounding distortion. Do this last.

The 470k resistor just before the phase inverter could be reduced to 220k. And the pot just before that could be a 500K.

I know you've indicated a wish to keep the output section stock, still I can't resist... What is the voltage at the EL84 pair cathodes now? Is the plate voltage 300-ish Volts as per the schematic? The bias point is set by the 100 ohm resistor and some experimenting here could make a difference. It depends on the stiffness of the power supply.

Many people prefer Hexfreds to do the rectifying. They generate less switching hash and seem to sound smoother.

Component type is a much debated topic. I have opinions, but will keep them to myself, for that is not a can of worms but an entire bait shop.

This may be more than you want to do but I'd try these things before cascading gain stages. Cool old amps voiced well are far more musical than tarted up preamp distortion generators (another opinion masquerading as fact!) And a well-voiced amp makes you love your stompboxes more. Plus, this amp is a good candidate for using an A/B/both switch at the inputs to give you footswitchable level and tone options.

Good luck, studio!

 

[Bryson]

Yeah. Maybe try the cathode bias cap thing on V1/V2 for more gain.

Shunting the grids of the phase inverter through a pot can yield a crude master volume thing, as can increasing the output tube cathode bias resistor value.

I'd leave it alone and use a gain boost pedal, and a power attenuator or L-Pad before the speaker.

If just for practicing at home though, i'd use a dist pedal and not waste the output tubes.

 

[Baltimore]

I would move the EQ. Add a 1uf (.68 is the usual marshall value) or so cathode bypass to the first stage, and add a preamp volume after that stage (you might need a 470k in series with the pot to not hit the second stage TOO hard, you can also try bypassing it with a 470pf ala marshall). Then change the cathode cap in the second stage to something bigger. Again, around 1uf, you can go bigger, but if you dig jcm800s I wouldn't go much bigger. The EQ will come after the second stage, as well as the master volume. If you still need more gain, change the master to a dual pot, post phase inverter. You could even try that first, and add the cathode caps. This will get some more gain, but I don't think it would sound too marshall-y with the EQ out front like that.

What I'm describing is similar to an orange AD15. The AD15 has a volume control post phase inverter, so you get some extra crunch out if it without the crazy volume. http://users.telenet.be/orangefg/orange_mods.htm for schematics.

That will get you in the ballpark. Having the EQ between stages will make it much more fender-y and will suck all of the gain out before it can clip the second triode. Moving the EQ will help a lot.

Kevin

 

[blandman74]

It looks like another trait of Marshall amps is the low output impedance before the tone stack, compared to fenders at least. In the 18W, it's about 23k, I think. In the JCM it's a cathode follower. I think that affects the sound in a noticeable way. Not much you can do about that, but I just thought I'd point that out. It has an effect on sound.

If you want more distortion, why not put in a switchable pair of diodes on one of the channels? It's cheesy according to some people's taste, but it's easy and comparatively non-destructive. You can experiment with a few different kinds of diodes, and even put in a put to "soften" or "skew" the clipping effect.

EDIT: I mean, "put in a POT" .... :roll:

 

[rodabod]

Thanks for the suggestions, guys.

One very kind forum member has offered to let me try an actual JCM800 preamp section which i can tag onto the input of the phase inverter section to see if i actually like that kind of sound (whatever it will be like...)

Out of interest, what do you guys make of this E*lectro H*armonix preamp schematic?

Looks Marshally to me, but as I said, I'm a novice when it comes to tubes. Can anyone explain why the cathode resisitor is relatively low on the first stage? I've seen this on a few schematics.

Cheers,

Roddy

 

[Gus]

As others posted maybe a distortion before the amp

I would try a Si NPN rangemaster type one I have a nice take on that circuit. ETK has tried it so has soundguy.

You could also try
http://www.diystompboxes.com/pedals/npn_boost.gif

FWIW Fun little boost the interaction with the hfe vs gain pot setting and input R. It was what could I do with one Si transistor that is different than other effect boosts. I had not seen the use of bootstrapping gain control interaction in a guitar effect schematic before the NPN boost.
Yes bootstrapping is old the univibe uses is in some stages. The LPB is a old guitar effect. But I have not seen a schematic for a guitar effect like this before.

Another mild boost with EQ try different Si transistors

http://www.diystompboxes.com/pedals/gusOverdrive.gif

 

[blandman74]

On the electro-harmonix schematic ...

I shouldn't judge if I haven't heard it, but it looks weird.

It uses 12AY7 tubes, not 12AX7s, so the gain will be about half for each stage.

On the plus side, the plate resistance is far lower, meaning the output impedance will be low before the tone stack, which makes it more Marshall-y.

C1 is way high, making it a fender-style bass roll-off. to make that input stage Marshall-y, you would make C1 a 2.2 uF for a low-end roll off of about 180 Hz. Caveat emptor on my math ....

EDITED to remove dumb mistakes. Still caveat emptor tho

 

[Johan]

before you do anything radical, why not just sacrifice one input. pot a 1M pot in, hook it up just like the "cut" pot on a Vox AC30, but without the cap...that way you have a, not perfect, but working, post P.I mastervolume and everything should be easy to reverse...
j

 

[kent]

Just reiterating what's already been said. After farting around with some different amp possibilities I came to the conclusion that to get decent preamp distortion the tone stack needs to be at the end before the PI and you do need the cathode follower. I can't put my finger on why the CF does the magic. Probably the afore mentioned impedance issues. The tone stack needs to be later because you've got carve that bottom out some so it doesn't get farty. I've tried modding closer to the Fender way without much success in the "grind" area. On the other hand... everything is subjective.

cheers & have fun modding,
kent

 

[CJ]

What the.........
.
.
.
.
Who the.........
.
.
.
Where the Hell you been!
:thumb:

 

[khstudio]

I have & built an 18w & it rips... but NOT at low volume.

If you want something that has killer BALLS at low volume... try the 1 watt.
it sounds like ACDC & kicks ass at low volume. It's also an easy build.

I'm not sure you'll get what your after modifying your amp.

 

[Emperor-TK]

Kevin, which 18W and 1W variation did you go for? I'm looking at doing a low wattage amp from scratch for my next project. Can you provide any details?

Thanks,
Chris

 

[kent]

Oh, CJ (sigh)...

I'm still a big time lurker. It's just that the kids keep me EXTRA busy these days. My DIY is mostly done vicariously through you guys.

... however, I do have a new amp on the bench right now :wink:

cheers all,
kent

 

[Gus]

http://www.ax84.com/dp.php?pg=legacyprojects&project_id=firefly

 

[khstudio]

[quote author="Emperor-TK"]Kevin, which 18W and 1W variation did you go for? I'm looking at doing a low wattage amp from scratch for my next project. Can you provide any details?

Thanks,
Chris[/quote]

The 18w is absolutely killer!
http://www.18watt.com

I built the TMB Version but if I had to do it again I'd just build the "STANDARD 18W"... I built one for a friend & I liked it a hair better than mine + I love the Tremolo. The TMB 's 2nd channel is set up like a PLEXI but there's 2 gain controls on that channel so you can overdrive the pre section (A LOT) which you may like for lower volume drive.
If you build it, I RECOMMEND the Heybor transformers!!!!!!

The 1watt is the FIREFLY, By Doug Hammond:
http://www.ax84.com/dp.php?pg=legacyprojects&project_id=firefly

http://www.diyguitarist.com/GuitarAmps/FireFly.htm

I tried to find Doug ORIGINAL site but it's gone! :sad:

He had 3 revisions of the Firefly & some good audio clips for each so you could hear the transition of each & choose for yourself... o well.

 

[rodabod]

Having looked at the amp last night, I'm leaning towards keeping the circuit as it is, but maybe changing some components for more gain on one channel. I added a 0.68uF cathode-bypass cap on the first stage which increased gain, but only enough for very slight overdrive at higher volumes. Sounded better to me though.

[quote author="dabo"]Tone stack revoicing is a good call too. I'd begin by reducing the .1uF to .047uF or .033 uF. (Check the calculator to see how it effects the bottom extension!) Controlling the low content is crucial to amp voicing (my opinion stated as though it were fact).[/quote]

I've tried the Duncan tone stack program and plotted the Wem's tone stack with modded values (but the same pots) against the standard Marshall tone stack:

(Red=Marshall, Blue=Modded Wem)

As you pointed out, my tone stack is very lossy due to the 250K pots, etc. and so I am losing around 12dB or so before hitting the second stage. It would be nice to change the pots, but they are PCB-mounted and I don't think I could get a replacement easily. Still, my modded tone stack follows the Marshall's curves reasonably.

One question I do have - I'm not sure about setting Zsource in the emulator. I don't see how a typical Zsource for a Marshall would be 1.3K - is this the typical output impedance for a cathode follower? I'm not sure what this value would be for the output of my Wem's first stage.

Since there is a volume pot after the second gain stage you may get away with increasing the 100K plate load resistor on the first stage to 120K or 150K and hit the second stage a bit harder. This can be a dangerous game if you get too gain greedy because the bias point will move and you really should also rethink the cathode resistor value and the bypass cap value too. You may end up with really bad sounding distortion. Then again, you may enjoy the flavor of this really bad sounding distortion. Do this last.

Since my tone stack is losing a fair bit of gain, would it be reasonable for me to go ahead and increase the plate load resistor. I can see the effect this has on a simple level, but not sure if have my head around the whole impact of changing this variable. I'm guessing I'll basically be losing headroom whilst increasing gain? Although it's perhaps not a good example to go by, I see Mesa use a 200K plate load resistor on the first stage of the Dual Rectumfriers. I'd like to maybe give this a try to see the effect, but I'm not sure about maintaining the bias point as mentioned.

I know you've indicated a wish to keep the output section stock, still I can't resist... What is the voltage at the EL84 pair cathodes now?

I'll see if I can measure this when I next have the amp open.


Cheers!

Roddy

 

[Gus]

try the fender value of 38K.